Why might trees in this forest support foliage with
high nutrient concentrations? One reason is that large
amounts of nutrients come into the system via inter-
ception of precipitation, especially wind-blown mist,
which may be facilitated by the interceptive capac-
ity of the abundant epiphytes (Nadkarni 1986b). Other
montane and lowland forests may not receive as much
direct atmospheric input in this form (Vitousek and
Sanford 1986). Second, the soils on which the forest
grows may be relatively fertile, in at least some ma-
cronutrients, as they are of recent volcanic origin (see
Chap. 2, Physical Environment). Mineralization rates
of nitrogen and phosphorus in the field on the forest
floor are also fairly high (Vance and Nadkarni 1990).
Nitrogen and phosphorus availability in our study
area, as indicated by high foliar concentrations and
only moderate retranslocation, appears comparable or
even higher than in other tropical montane forests.
Nutrient use efficiency (calculated as biomass/nutri-
ent return) is low relative to other tropical forest and
other montane forests (Table 9.20). The Monteverde
forest may be less limited by nutrients than in other
cloud forests. Cloud forests as a vegetation type are
diverse with respect to their nutrient cycling regimes.
Epiphytic litter. In many tropical moist forests, live
epiphytes and their associated dead organic matter on
branches and trunks constitute up to 45% of the fo-
liar mineral capital (Pocs 1980, Nadkarni 1983). Nu-
trients from live and dead ephiphytic organic matter
(EM) are released into the nutrient cycles of terrestri-
ally rooted vegetation by three pathways: (1) epiphyte
mats on host tree branches and trunks are permeated
by host tree canopy roots (Nadkarni 1981), (2) epi-
phyte mats are leached by precipitation and the nu-
trients are transferred to the forest floor via stemflow
and throughfall, and (c) EM falls to the forest floor and
decomposes. Processes that cause EM to fall to the
forest floor include senescence, wind, disruption by
birds and mammals, and the falling of supporting
branches and whole trees. The contribution of the
epiphyte community to nutrient transfer in tropical
forests is poorly understood. The few reports in which
epiphyte litterfall has been reported in tropical for-
ests have been anecdotal or based on small collectors
designed to trap tree fine litter (Tanner 1980a, Songwe
et al. 1988), rendering estimates of the total input of
nutrients to the forest floor inaccurate in forests where
epiphytes are a substantial canopy component.
We quantified the dynamics of fallen EM in our
study area, specifically (1) EM standing crop biom-
ass, composition, and nutrient pools on the forest
floor; (2) input of EM biomass, composition, and nu-
trients to the forest floor; (3) rates of EM biomass and
nutrient turnover (Nadkarni and Matelson 1992b); we
also made preliminary estimates of short-term decom-
position. Samples of fallen EM were separated to vas-
cular plants, bryophytes, and dead organic matter. To
collect the large discrete pieces of EM litterfall that
would not fit into standard litterfall traps, we col-
lected newly fallen EM from twenty 5 x 5 m cleared
plots twice per month. To measure EM that fell in the
form of smaller pieces, we used the tree fine litterfall
collectors on the forest floor (Nadkarni and Matelson
1992b).
In 1988, the mean biomass of EM standing crop was
50 g/m^2 ; in 1990, the mean was 27 g/m^2. There was a
great deal of spatial variation in the amount of stand-
ing crop (range 330-8190 kg/ha). Composition of
standing crop was dominated by dead organic matter
(58%), followed by bryophytes (22%), and vascular
plants (20%). The nutrient pool of standing crop on
the forest floor (Table 9.21) was calculated by multi-
plying the biomass of each component by the nutri-
ent concentration of that component.
The biomass of EM input from plots during the
study period was 350 kg/ha/yr. Input was highly vari-
able spatially, with biomass from individual plots
ranging between 0 and 232 g/m^2 per collection period.
Standard deviations for a given collection interval
were between 6% and 360% of the mean. EM input
measured with the fine litter collectors was 140 kg/
ha/yr, which is equivalent to 2% of the total terres-
trially rooted fine litter (Nadkarni and Matelson
1992a). This bryophyte component was added to plot
collection input, for a mean total annual EM input of
0.5 tons/ha.
Input of EM was temporally sporadic; greater
amounts fell in 1988-1989 than in 1989-1990 (Fig.
9.13). The highest values of EM litterfall occurred
during 1988 windstorms, which were the most severe
recorded in the past 15 years (J. Campbell, pers. comm.).
Of 1234 collections of individual plots, 99% con-
tained bryophytes, 62% contained vascular plants, and
56% contained dead organic matter. Composition of
EM input to the forest floor on a dry weight basis was
bryophytes, 76%; dead organic matter, 13%; and vas-
cular plants, 11%. Individual collections of extremely
large samples were infrequent; only 26 (2%) of all
collections exceeded 10 g/m^2 per collection. These
sporadic pulses of large EM comprised 53% of the
total EM input. A continual but unpredictable input
of small amounts of material fell throughout the year
(Fig. 9.13). Nutrient input via EM litter (Table 9.21)
was calculated by multiplying the nutrient concen-
tration (Table 9.22) for each component at each col-
lection period by the mean biomass of that compo-
nent for that collection period.
In our study site, the biomass of EM litterfall (0.5
tons/ha/yr) is equivalent to 5-10% of the total fine332 Ecosystem Ecology and Forest Dynamics